OpenMp根据变量设置并行循环的线程数 [英] OpenMp set number of threads for parallel loop depending on variable

问题描述

有没有一种方法可以基于变量的值在OpenMP并行区域中设置线程数?最初，整个应用程序的线程数= nofCores.在我的AMD FX 8350上，nofCores = 8.对于此区域，如果变量为3，那么我只需要3个线程.如果variable> cores，则线程数应保持等于nofCores.

Is there a way to set number of threads in OpenMP parallel for region based on the value of a variable? Initially for the whole application number of threads = nofCores. On my AMD FX 8350, nofCores =8. For this area if the variable is 3 then I only need 3 threads. If variable>cores then number of threads should remain equal to nofCores.

我不想为所有应用程序全局设置线程数.仅用于此特定的并行循环.

I do not want to set the number of threads globally for all the application. Just for this specific parallel loop.

很抱歉，如果这是一个幼稚的问题，但我是OpenMP的新手.

Sorry if this is a naive question, but I am a newbie in OpenMP.

推荐答案

当然..只需将其添加到您的parallel for指令即可:

Sure .. just tack this on to your parallel for directive:

#pragma parallel for num_threads(variable)
for( ... )

这篇关于OpenMp根据变量设置并行循环的线程数的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！